j River Deposits of the Spring River Valley .— Hershey. 37 
3. They indicate a possible explanation of the coarse par¬ 
ticles of sand found in otherwise very fine deposits.* 
My own observations and experiments confirm conclusions 
1 and 3. I find, in addition : 
1. That sand grains will float in perfectly still watet for 
an indefinite time. 
2. That the grains which float are not necessarily siliceous. 
That flakes of mica, fragments of marble, bituminous shale, 
etc., also float and that some of them, the marble and the 
bituminous shale, for example, are unusually buoyant. 
3. That the property of floating is not confined to the 
$and of any particular locality but depends to a considerable 
extent upon the angularity, i. e. the shape of the grains. 
4. That whether sand will float or not depends, also, upon 
the mode of launching. Whether it be by ripple waves, as 
stated by Mr. Graham, or by under-mining, it must be gently 
done, for should the grains be plunged into the water with 
sufficient force to completely immerse them they will immed¬ 
iately sink. 
5. That the natural conditions necessary to the floating 
of sand in rivers are somewhat unusual, depending, in the 
case of the Llano, upon a flood without local rains and, in 
that of the Connecticut, upon the manner in which certain 
waves strike a sand-bar. It is quite possible, however, that 
floating sand is much more common than is ordinarily sup¬ 
posed. 
6. That the physical explanation of the problem is com¬ 
plex rather than simple, and at best unsatisfactory in several 
important particulars, and that with the advance of molec¬ 
ular physics we ma}^ hope for a better understanding of 
what we now, for convenience, term “ superficial viscosity ” 
and “ capillary attraction.” 
School of Geology , University of Texas. JVov ., 1895. 
ANCIENT RIVER DEPOSITS OF THE SPRING RIVER 
VALLEY IN KANSAS. 
By Oscar H. Hershey, Freeport, Ill. 
In the February (1895) number of the Am. Geologist, Mr. 
Arthur Winslow mentions certain gravel deposits along Spring 
*Amer. Jour. Sc., Ill, vol. xl, p. 476. 
